2 Compound Processing Characteristics and Testing2 Compound Processing Characteristics and Testing...

2 Compound Processing Characteristicsand Testing

John S. Dick

2.1 Introduction

The processing of a rubber formulation is a very important aspect of rubber compounding.

Also, it is important to understand the rubber properties needed for good processability and

the tests performed to measure these properties.

2.2 Manufacturing Process

Uncured rubber, whether natural or synthetic, behaves as a viscoelastic ¯uid during

mixing. Under processing conditions, various rubber chemicals, ®llers, and other additives

can be added and mixed into the rubber to form an uncured ``rubber compound.'' These

compounding ingredients are generally added to the rubber through one of two basic types

of mixers, the two roll mill or the internal mixer.

2.2.1 Two Roll Mill

The two roll mill in Fig. 2.1 consists of two horizontal, parallel, heavy metal rolls which

can be jacketed with steam and water to control temperature. These rolls turn towards each

other with a pre-set, adjustable gap or nip to allow the rubber to pass through to achieve

high-shear mixing. The back roll usually turns at a faster surface speed than the front roll;

this difference increases the shear forces. The difference in roll speeds is called the friction

ratio. The rubber generally forms a ``band'' around the front roll. Mill mixing is the oldest

method of rubber mixing, dating back to the very beginning of the rubber industry;

however, it is a relatively slow method and its batch size is limited. Internal mixers

overcome these problems.

2.2.2 Internal Mixers

Internal mixers were ®rst developed by Fernley H. Banbury in 1916. Today, these internal

mixers are commonly used because they are much more productive than two roll mills.

Internal mixers consist of two rotors or blades turning toward each other in an enclosed

metal cavity. These rotors can be either tangential or intermeshing in arrangement

(Fig. 2.2). The cavity is open to a loading chute through which rubber, ®llers, and

various chemicals are placed. Upon completion of the mix cycle, the mixed rubber stock is

discharged through a door in the bottom of the mixer.

Mixing time is determined by the shape and size of the rotors, the rotor speed, and

horsepower of the motor turning them. The rotors generally turn at a high friction ratio.

Some internal mixers can handle batches in excess of 1000 pounds (greater than 455 kg),

and in some cases, can completely mix a compound in less than two minutes. Of course,

with so much energy being absorbed by the rubber stock, the batch temperature can rise

well above 120 8C (250 8F) before it is dumped and cooled. The temperature rise that

results from viscous heating of the rubber compound often means the compounds must

pass through the internal mixer more than once to disperse ®llers and other compounding

Figure 2.1 Two roll mill.

Figure 2.2 Tangential vs. intermeshing rotors for internal mixers.

18 J.S. Dick [Refs. on p. 44]

ingredients. Batches are usually dumped from an internal mixer onto a mill where they

may be further worked while being cooled. Sometimes additional compounding ingre-

dients, such as curatives, are added at this point.

2.2.3 Further Downstream Processing

After the rubber stock is mixed, it may be remilled and fed into either a calender or

extruder. A three or four roll calender (Fig. 2.3) is generally used when uncured rubber is

applied to a textile fabric or steel cord as a coating. Calenders are also used to form sheets

of rubber. Extruders (Fig. 2.4) are used when the uncured rubber stock is to be shaped into

a tire tread, a belt cover, or a hose tube, for example. A screw rotating in a cylinder

transports the rubber compound from the feed port to the die of the extruder. During this

process, additional heat and work histories usually affect the viscoelastic behavior of the

compound and how it extrudes through the die. Extruders with lower screw length to screw

diameter (L/D) ratios are considered ``hot feed'' extruders while ones with high ratios are

considered ``cold feed'' extruders.

After calendering or extrusion, the uncured rubber parts may be brought together in a

building step. For example, uncured tire components are put together by hand or

automatically on a tire building machine where the calendered carcass plies are placed

over the innerliner; then, the belt and sidewalls over the carcass, etc. In a similar manner,

conveyor belts are constructed on large building tables on which each calendered ply is

Figure 2.3 Three and four roll calenders.

2. Compound Processing Characteristics and Testing 19

manually placed over another, followed by the extruded cover stock. Some hose may be

hand built as well. Other rubber products may require that the mixed compound be

preshaped and molded.

2.2.4 Curing Process

After the shaping and/or construction of the uncured (``green'') rubber product, it goes

through the cure process. With tires, the uncured constructed assembly is placed in a

special tire press, such as a Bag-O-Matic press, where it is shaped in a tire mold and cured

at elevated temperatures under pressure. Other products, such as hose and rollers, are

sometimes cured in autoclaves with steam pressure. Many automotive rubber parts are also

cured through compression molding, transfer molding, or injection molding. Rubber cure

temperatures can generally range from as low as 100 to more than 200 8C. The higher curetemperatures mean shorter cure cycles. Many rubber products are therefore cured at higher

temperature ranges for greater productivity, provided that reversion or thermal degradation

is not a problem at these higher temperatures. Most rubber products are cured under

pressure as well to avoid gas formation and porosity.

2.2.5 Factory Problems

If rubber compound processability characteristics are not controlled for the different stages

of the manufacturing process, various factory quality problems can result. Poor control can

Figure 2.4 Rubber mixing extruder.


result in higher scrap rates, higher internal and external failure costs, and lost plant

productivity, all of which can command a great deal of the rubber compounder's attention

as daily plant problems. Some examples [1] of these problems are given in Table 2.1.

Lord Kelvin once said that ``applying numbers to a process makes for the beginnings

of a science.'' If rubber processability characteristics can be measured effectively, then the

rubber compounder can ®nd the cause of many of the factory problems given in Table 2.1.

Various test methods help determine the cause of many problems and aid in the design of

new compounds and/or better processing conditions to reduce factory problems.

2.3 Processability Characteristics and Measurements

The following are rubber compound processability characteristics that determine how well

a given compound processes:

Table 2.1 Examples of Factory Problems

MixingLong black incorporation times (BIT)Lower ®ller dispersionHigher compound viscosityCrumbly batchesLumpy stocksSlow mixing

MillingBack rollingBaggingPoor mill releaseThickness control

ExtrusionRough surface (not smooth)Poor dimensional stabilityUncontrolled extrudate shrinkageHigh die swellLow extrusion ratePin holesScorch

CalenderingScorchBare spots, holesPoor dimensional stability, width, thicknessHeat blistersTrapped airCalender release (from rolls)Rubber fabric penetrationLiner release

Stock bin storageScorchGreen strengthSticky slabs

Compression/transfer/injection moldingPoor appearanceShrinkageCured hardnessPorosityMold releaseMold fouling

Autoclave curesAppearanceMandrel releaseShrinkage

Second step tire buildingPoor green strengthBlow outs

Tire moldingUnder®llsPoor appearanceUniformityPorosityMold foulingVent problems


1. Viscosity

2. Shear Thinning

3. Elasticity (or V/E Ratio)

4. Time to Scorch

5. Cure Rate

6. Ultimate State of Cure

7. Reversion Resistance

8. Green Strength

9. Tack

10. Stickiness

11. Dispersion

12. Storage Stability

13. Mis-Compounding (Compound Variation)

14. Cellular Rubber Blow Reaction

2.3.1 Viscosity

Viscosity is the resistance of a ¯uid, such as rubber, to ¯ow under stress. Mathematically,

viscosity (Z) is shear stress divided by shear rate as shown below in Eq. (2.1).

Z � shear stress

shear rate�2:1�

Viscosity is very dependent on temperature; at higher temperatures materials are less

viscous. The viscosity of rubber can be measured by four methods:

1. rotational viscometers

2. capillary rheometers

3. oscillating rheometers

4. compression plastimeters

2.3.1.1 Rotational Viscometers

By far, the most commonly used rotational viscometer in the rubber industry is the

Mooney viscometer. Melvin Mooney of U.S. Rubber Co. developed this instrument in the

1930s [2]. Since then, it has become one of the most widely used test methods in the

industry. It is used for testing both raw rubber and mixed stocks.

This method is described in detail in ASTM D1646 or International Standard ISO 289,

Part 1. Two precut rubber test pieces with a combined volume of 25 cm3 are placed into a

two-part compression cavity mold. With the dies closed, a sealed, pressurized cavity is

formed, in which a special rotor is imbedded in the rubber. This rotor and the dies are

grooved to help prevent the rubber from slipping at the rotor or die interface while the

rotor is turning.

Usually there is a pre-heat time after the dies are closed to allow the rubber to approach

the set temperature of the instrument. Then the test speci®cation calls for the rotor to turn at

two revolutions per minute (2 rpm) for a speci®ed time period. The instrument records


viscosity in Mooney Units (MU), which are arbitrary units based on torque. Generally, the

measured viscosity of the rubber under test decreases with running time because of the

thixotropic effects of the rubber tested. However, depending on the type of rubber and the test

temperature, the rate of decrease in the measured Mooney viscosity with time should slow

down greatly (Fig. 2.5). Usually, the Mooney viscosity test is performed with a one minute

preheat and a run time of either four or eight minutes. The ®nal Mooney viscosity value is

reported as the lowest value recorded in the last 30 seconds of the test.

As an example, a Mooney viscosity value may be reported as 55ML (1� 4). This is

in accordance with the convention recommended by the ASTM standard cited above. The

term ``ML'' indicates a large standard rotor was used. ``55'' represents the measured

Mooney viscosity value reported for the speci®ed conditions of the test. The ``1''

represents the preheat time before the rotor starts to turn. The ``4'' indicates the running

time of actual rotation of the rotor before the ®nal Mooney viscosity measurement is

made.

Mooney viscosity crudely relates to the average molecular weight of a raw rubber and

to the state-of-mix or quality-of-mix for a masterbatch or ®nal uncured rubber stock. If the

Mooney viscometer with a large rotor is measuring values well above 80MU, it can be

somewhat insensitive to subtle differences among raw polymers or mixed stocks. When

the rubber specimen is too tough, slippage and tearing may occur. One method to avoid

this is to simply run the test at a higher temperature. For example, many EPDM polymers

do not test well at 100 8C. Instead, they are tested at 125 8C. However, a higher

temperature can be a problem for rubber compounds containing curatives that are

``scorchy.'' Also, if a rubber has a high Mooney viscosity from the large rotor (ML),

then repeating the test with a small Mooney rotor (MS) should be considered. One basic

problem with the Mooney viscometer is that it measures viscosity at a low shear rate of

only 1 sÿ1, which is far lower than many rubber manufacturing processes.

2.3.1.2 Capillary Rheometer

The capillary rheometer measures the viscosity of mixed rubber stocks at relatively high

shear rates. ASTM D5099 describes this method for rubber testing. There is no ISO

Figure 2.5 Mooney viscosity curve of raw rubber.


International Standard for the capillary rheometer's use with rubber. The ASTM method

consists of placing (or packing) cut pieces of a rubber sample in a heated barrel. Then, a

special piston pushes the rubber out of the barrel through an ori®ce of a special die (with a

given capillary length to ori®ce diameter ratio or L/D) to form an extrudate. The apparent

shear rate is determined from the speed of the piston traveling in the barrel (the ram

speed). The apparent shear stress is determined from the resulting barrel pressure measured

by a transducer (Fig 2.6). From Eq. (2.1), the apparent viscosity Zapp can be calculated.

This apparent viscosity can be converted to ``true'' viscosity by applying the Rabinowitsch

correction to obtain the ``true'' shear rate and the Bagley correction to obtain the ``true''

shear stress [3,4].

The main advantage of the capillary rheometer for measuring rubber viscosity is the

wide shear rate range it can apply to the rubber specimen. Many capillary rheometers can

measure viscosity shear rates at over 1000 sÿ1. The disadvantages of these rheometers are

that they are dif®cult to operate, require more time to run a single test, and require

extensive time to clean the barrel and set up for the next test.

Figure 2.6 Capillary rheometer.


2.3.1.3 Oscillating Rheometers

An instrument called the Rubber Process Analyzer (RPA) by Alpha Technologies [5]

measures dynamic viscosity through the application of a sinusoidal strain to an uncured

rubber specimen molded in a sealed, pressurized cavity. ASTM D6204 describes this

unique method for measuring processability. Figure 2.7 shows the instrument used for this

method. Figure 2.8 shows a sinusoidal strain being applied to the rubber test specimen.

The complex torque response (S�) is observed to be out-of-phase with the applied strain

because of the viscoelastic nature of the rubber being tested. The phase angle d quanti®es

this out-of-phase response.

Figure 2.7 RPA 20001 rubber process analyzer.

Figure 2.8 Applied sinusoidal strain and resulting stress response.


From the complex torque S� response and the phase angle d, the elastic torque S0 (in-phase with the applied strain) and the viscous torque S00 (908 out-of-phase with the applied

strain) can be derived (Fig. 2.8). The elastic response S0 is a function of the amplitude of the

applied strain while the viscous torque is a function of the rate of change of the applied strain.

The storage shear modulus (G0) is calculated as follows:

G0 � k � S0=strain �2:2�The loss shear modulus (G00) is calculated as follows:

G00 � k � S00=strain �2:3�where k is a constant that takes into account the unique geometry of the die cavity.

The complex shear modulus (G�) is equal to the square root of the sum of G0 squaredand G00 squared.

G� � ��G0�2 � �G00�2�1=2 �2:4�The complex dynamic viscosity Z� is calculated as follows:

Z� � G�=o �2:5�where o is the frequency of the sinusoidal strain in radians/second.

The real dynamic viscosity Z0 is also calculated in the following manner:

Z0 � G00=o �2:6�The complex dynamic viscosity Z� from the Rubber Process Analyzer is analogous to the

apparent viscosity (Zapp) from the capillary rheometer, while the real dynamic viscosity Z0

is analogous to the ``corrected'' viscosity from the capillary rheometer.

Cox and Merz in the early 1950s published the following empirical relationship between

capillary rheometer viscosityZapp measured under conditions of steady shear rate and dynamic

complex viscosity Z�, which is measured through sinusoidal deformation (and constantly

changing shear rates) applied by a dynamic mechanical rheological tester [6]:

Zapp� _g� � Z��o�jo� _g �2:7�� Zapp is the apparent (uncorrected) capillary rheometer viscosity at a steady shear rate

(in sÿ1)� Z� is complex dynamic viscosity measured at an oscillatory frequency of o (in radians

per second).

This empirical relationship sometimes works, but not every time [7,8]. Sometimes the

success of this empirical relationship is based on the nature and concentration of the

reinforcing ®llers used.

The important advantages of the Rubber Process Analyzer in measuring rubber

viscosity over the other methods discussed are its versatility in measuring viscosity at

both low and high shear rates, ease of use, and excellent repeatability.

2.3.1.4 Compression Plastimeters

While viscosity is de®ned as the resistance to plastic deformation, the term plasticity refers

to the ``ease of deformation'' for a rubber specimen. In a way ``plasticity'' and


``viscosity'' de®ne the same property, but have the opposite meaning. A plastimeter

measures the plasticity of an uncured rubber specimen. Plastimeters are very simple, crude

methods for measuring the ¯ow of a rubber sample. The main problem with most

plastimeters is that they operate at extremely low shear rate ranges of only 0.0025 to

1 sÿ1, much lower than even the Mooney viscometer [9]. To reinforce this point, the

Rubber Process Analyzer can be correlated to these methods but only when very low

frequencies (very low shear rates) are applied to the rubber specimen [10]. The principle of

plastimeter procedures is basically to measure the deformation of a cylindrically cut,

uncured rubber specimen after it has been subjected to a constant compressive force

between two parallel plates for a speci®ed time period at a speci®ed test temperature.

The initial part of ASTM D 3194 (the standard describing the Plasticity Retention

Index for natural rubber) and International Standard ISO 2007 describe the use of the

Wallace rapid plastimeter method. The Williams plastometer parallel plate method is

described by ASTM D926 and ISO 7323.

2.3.2 Shear Thinning

Shear thinning is the characteristic of non-Newtonian ¯uids (such as rubber compounds) to

decrease in measured viscosity with an increase in applied shear rate. Not only is it

important to measure the compound viscosity, it is also important to know how rapidly

viscosity decreases with increasing shear rate. All rubber compounds are non-Newtonian

in their ¯ow characteristics and their viscosity usually decreases according to the power

law model. If the log of viscosity is plotted against the log of shear rate, a straight line

usually results. Rubber compounds with different reinforcing ®ller systems have different

log-log slopes. These different slopes can be quite important because compounds are

commonly processed at different shear rates. As shown in Fig. 2.9, the ordinal relationship

of Compound 1 vs. Compound 2, for example, can change between a low shear rate to

high shear rate. So, Compound 2 might have higher viscosity than Compound 1 in an

Figure 2.9 Comparison of two compounds with different shear thinning pro®les from capillary rheometer

viscosity measurements with increasing applied shear rates.


injection molding operation, but Compound 2 might have lower viscosity than Compound 1

in the mold after injection.

Two effective methods of measuring shear thinning behavior are the capillary

rheometer and oscillating rheometer.

2.3.2.1 Shear Thinning by Capillary Rheometer

The capillary rheometer can apply a wide range of shear rates. As discussed earlier,

according to ASTM D5099, the piston can be programmed to travel in the barrel at a series

of faster and faster speeds which result in higher and higher shear rates. Figure 2.10 shows

the log-log plot for a capillary rheometer. The applied shear rate for a capillary rheometer

can be more than 1000 sÿ1.

2.3.2.2 Shear Thinning by Oscillating Rheometer

As discussed earlier, the Rubber Process Analyzer (which is de®ned in ASTM D6204)

increases shear rate by increasing the frequency of the sinusoidal strain. Figure 2.11 shows

a log-log plot of the complex dynamic viscosity vs. shear rate (frequency in radians/

second) for the same set of compounds shown earlier in the ®gure for capillary rheometer

measurements. In fact, Fig. 2.12 shows how the power law slopes from these measure-

ments for capillary rheometer vs. Rubber Process Analyzer agree quite well [11].

Shear rate with the Rubber Process Analyzer can also be increased by using a higher

applied strain. Sometimes applying higher strains to achieve higher shear rates may have

some advantages over higher frequencies because these higher strains may more

effectively destroy carbon black aggregate-aggregate networks which form in the rubber

compound while in storage. Figure 2.13 shows the very good correlation achieved between

a capillary rheometer shear stress values measured at 100 sÿ1 and Rubber Process Analyzer

G0 values at very high applied strain [12].

Figure 2.10 Log-log plot of MPT capillary rheometer.


The Rubber Process Analyzer is very simple to operate and versatile in applying

different shear rates by either varying the frequency or varying the strain. It is a unique

instrument because it can apply very high strains to an uncured rubber specimen in a

sealed pressurized cavity and achieve very repeatable results. Also, the Rubber Process

Analyzer has an advantage over the capillary rheometer in that it can more effectively test

raw rubber samples for shear thinning behavior and achieve better repeatability.

2.3.3 Elasticity

Elasticity is that material property that conforms to Hooke's Law. A purely elastic

material, such as some metals at low strains, conforms perfectly to Eq. (2.8) below [13]:

s � Eg �2:8�

Figure 2.11 Log-log plot of RPA dynamic complex viscosity Z� vs. rad./s (frequency sweep).

Figure 2.12 Comparison of power law slopes for shear thinning as measured with a capillary rheometer

vs. an oscillating rheometer (RPA).


where:

� s � stress, or force per unit area

� g � strain (displacement) as measured from change in length

� E � the static modulus of elasticity

With a perfectly elastic material, the rate of applied deformation has no effect on the

measured stress response. Of course, rubber is not perfectly elastic even in the cured state.

Rubber is viscoelastic, possessing both viscous and elastic qualities in both the uncured

and cured states. However, the ratio of the viscous quality to the elastic quality (the V/E

ratio) decreases greatly when a rubber compound is cured.

Uncured rubber possesses an elastic quality mainly as a result of chain entangle-

ments. Uncured rubber with a high elastic quality has what is commonly called nerve.

Rubber with high ``nerve'' (high elasticity) resists processing. It is not uncommon for two

rubbers to have the same viscosity, but one is ``nervier'' (has more elasticity) than the

other. This elastic quality difference affects how well the rubber processes. This higher

elastic quality affects how well a rubber mixes, how well it incorporates ®llers, the length

of a mix cycle, and the viscoelastic properties imparted to the ®nal batch. For example, a

mixed stock with a higher elastic quality may have greater die swell, poorer dimensional

stability during a downstream extrusion process, or mold differently from a stock with

higher nerve.

The elasticity of rubber can be measured by ®ve test methods:

1. Mooney stress relaxation

2. oscillating rheometer

3. capillary rheometer die swell

Figure 2.13 MPT capillary rheometer shear stress at 100 sÿ1 vs. RPA S0 at 908 arc strain.


4. compression plastimeter elastic recovery

5. direct shrinkage measurements.

2.3.3.1 Mooney Stress Relaxation

In 1996, a new Part B for measuring stress relaxation was added to ASTM D1646, the

Mooney Viscosity Standard. This new method is widely used in the rubber industry

because it can measure Mooney viscosity as well as stress relaxation decay rates. Stress

relaxation tests are run on the same specimen immediately after completing viscosity

measurements. For example, an ML 1� 4 test (large rotor, 1 minute preheat, 4 minutes

run) can be performed on a raw rubber sample followed by a two minute stress relaxation

at the end of the test. The total test time is seven minutes. Mooney stress relaxation can be

performed automatically after the ``®nal'' Mooney viscosity measurements by very

rapidly stopping the rotation of the rotor and measuring the power law decay of the

Mooney viscosity output with time. Mathematically, this power law decay is described in

Eq. (2.9).

M � ktÿa �2:9�

where

� M is the torque value in Mooney units

� k is the torque value at 1 second

� t is the time in seconds

� a is the rate of relaxation (the slope of the relaxation function)

In a log-log plot, this expression takes the following form:

logM � ÿa log t � log k �2:10�

The slope a is commonly used as a measure of the stress relaxation for raw rubber and

rubber compounds. Figure 2.14 illustrates how this slope a can be used to quickly compare

the elasticity of two EPDM polymers. These two EPDMs have the same Mooney viscosity

Figure 2.14 Mooney stress relaxation for two EPDM polymers using the MV20001 Mooney viscometer.


but different elasticity. A steeper slope (a faster rate of decay) indicates that the polymer

has a higher V/E ratio and a lower elastic quality than the polymer shown with a ¯atter

slope. Therefore, these two polymers are likely to process differently [14].

2.3.3.2 Elasticity by Oscillating Rheometer

ASTM D6204, which describes the Rubber Process Analyzer (RPA), can provide a

procedure to measure the elasticity of rubber. As described in Section 2.3.1.3, the storage

modulus G0 and loss modulus G00 are measured by this technique from the testing of raw

rubber as well as uncured mixed stocks. G0 is a direct measure of elasticity. A higher G0

response means a higher elastic quality at a de®ned temperature, frequency and strain.

Dividing G00 by G0 calculates a parameter known as tan d as shown in Eq. (2.11).

tan d � G00=G0 �2:11�Tan d is the tangent of the phase angle discussed in Section 2.3.1.3 (see also Fig. 3.5). A

higher uncured tan d for a raw elastomer or mixed stock implies a higher V/E ratio, which

is also analogous to a steeper slope from a stress relaxation test. The tan d, as a

processability parameter, has been found to be at least twice as sensitive to real

differences in rubber processability than the a slope from the Mooney Stress Relaxation

test discussed earlier [15]. Also, an RPA tan d as a parameter is more reliable and can be

measured at a wider range of shear rates than the a slope. Therefore, tan d and G0 are bothvery effective processability measurements. An example of comparing viscoelastic

properties with the RPA is shown in Fig. 2.15. Here, two different sources of SBR

1006 have almost exactly the same Mooney viscosity but have quite different tan d values.Other studies have shown similar differences [16,17].

Figure 2.15 Uncured tan d response from the RPA for two SBR 1006 polymers with the same Mooney

viscosity.


2.3.3.3 Capillary Rheometer Die Swell

As discussed earlier, ASTM D5099 describes the use of the capillary rheometer for testing

the viscosity of rubber compounds at different shear rates. When the rubber compound is

extruded through a standard die ori®ce, it swells from its elastic memory. A rubber

compound with high elasticity usually displays greater die swell. ASTM Test Method

D5099 does not describe a method for measuring this die swell. However, a capillary

rheometer, such as the Monsanto Processability Tester, can measure die swell directly

using a special optical detector. This method has a disadvantage in that its statistical test

sensitivity and repeatability are not nearly as good as the RPA method described earlier.

Also, the capillary rheometer is not normally effective for testing raw rubber.

2.3.3.4 Compression Plastimeter Elastic Recovery

As discussed earlier, the principle of these plastimeter procedures is to measure the

deformation of a cylindrically cut, uncured rubber specimen after it is subjected to a

constant compression force between two parallel plates for a speci®ed time period at a

speci®ed test temperature. This relates to plasticity, or the viscous component, as discussed

earlier. However, these plastimeter methods usually have an additional procedure for

elastic recovery, which is measured dimensionally in a given time period after the

deformational load has been removed. This elastic recovery is supposed to relate to the

elasticity of the rubber specimen. Some problems with these procedures in measuring

elasticity are that they are performed at low strain, low shear rate, and at a temperature that

is usually much lower than the factory processing temperature. Because of these

conditions, the data may not be relevant to plant processes. Variations in sample

preparation also greatly affect test results. Recovery data can be quite variable.

2.3.3.5 Direct Shrinkage Measurements

Another way of measuring nerve is to directly measure percent shrinkage. An example of

this is presented in ASTM D1917, which is the standard method for measuring shrinkage

of raw and compounded hot-polymerized SBR. Rubber strips from a mill are prepared and

placed on a speci®ed type of low-friction surface. The strips are exposed to elevated

temperatures for speci®ed lengths of time, and ®nally cooled, dimensionally measured, and

their percent shrinkage calculated. Even though this is an old method, no precision

statement has been published. This method is not widely used and some believe it is not

very repeatable.

2.3.4 Time to Scorch

The time to scorch is the time required at a speci®ed temperature (or heat history) for a

rubber compound to form incipient crosslinks. When a scorch point is reached after a

compound is exposed to a given heat history from factory processing, the compound

cannot be processed further by milling, extruding, calendering, etc. Therefore, scorch


measurement is very important in determining whether a given rubber compound can be

processed in a particular operation. The scorch time can be measured by:

1. Rotational viscometer

2. Oscillating rheometer

3. Capillary rheometer

2.3.4.1 Scorch by Rotational Viscometer

The Mooney Viscometer has been used to measure scorch since the 1930s. It was the ®rst

instrument method used to measure scorch safety of a mixed stock. ASTM D1646

and International Standard ISO 289, Part 2 describe how Mooney scorch time is

reported. Usually, the Mooney viscometer is set at a temperature higher than 100 8C,such as 121 8C or 135 8C, for example. From a practical viewpoint, some rubber

technologists consider the best scorch test temperature is selected when the compound

routinely reaches scorch within 10 to 20 minutes. However, others feel the temperature for

the Mooney scorch test should be close to the normal temperature of the factory process in

question.

Mooney scorch is generally reported for a large rotor as the time required for the

viscosity to rise ®ve Mooney units above the minimum viscosity (referred to as t5).

However, when the small rotor is used, the scorch time is reported as the time required for

viscosity to rise three Mooney units above the minimum (referred to as t3). It should be

noted that the Mooney rotor is unheated, which means that Mooney Scorch values are not

true isothermal measurements.

2.3.4.2 Scorch by Oscillating Rheometer

The oscillating disk rheometer (ODR) was introduced in 1963 and is considered a great

improvement over the Mooney viscometer because the ODR measures not only scorch, but

also cure rate and state of cure [18]. The ODR method is described in ASTM D2084 and in

International Standard ISO 3417. Unlike the Mooney viscometer, which continuously

rotates its rotor, the ODR oscillates its biconical disk (rotor) sinusoidally. This gives the

ODR an advantage in that it can measure the complete cure curve.

ODR scorch time is usually de®ned as the time until one torque unit rise above the

minimum is achieved (tS1) when 18 arc strain is applied, or time until two torque units rise

above the minimum is achieved (tS2) when 38 or 58 arc strain is applied. Also, through the

computer software used today with ODRs, it is relatively easy to calculate time to 10%

state-of-cure (tC10). This cure time tC10 also relates to scorch. The parameter tC10 may

have an advantage over tS1 because it is not based arbitrarily on a one torque unit rise

(which can be either dNm or in-lb torque units). Because the ODR measures the entire

cure curve, it is usually performed at a higher temperature than the Mooney scorch test

discussed earlier. As a result, the ODR may not be as sensitive to scorch differences as the

Mooney Scorch test [19].

Even though thousands of ODRs have been used worldwide, the ODR itself has a

design ¯aw which involves the use of the ODR disk itself. This is why the rubber industry


is shifting away from the ODR to rotorless curemeter designs (discussed next). The

problems associated with the disk rotor are as follows [20]:

1. The ODR disk itself functions as a ``heat sink,'' preventing the rubber specimen from

reaching the set temperature quickly. Because the rotor is not heated, this means that

the ODR scorch values are not a true isothermal measurement.

2. The ODR torque signals must be measured through the shaft of the rotor. This design

results in poor signal-to-noise response resulting from the friction associated with the

rotor.

3. This same friction associated with the ODR rotor prevents true dynamic properties

from being measured during cure.

4. After an ODR test, the cured specimen must be physically pried off the rotor. Because

of the rotor shaft, it is not easy to use a barrier ®lm. This also means that the ODR dies

are easily fouled and require constant cleaning.

5. The ODR is very dif®cult to automate because the cured specimen must be removed

from the rotor.

These problems were greatly reduced with the introduction of new rotorless

curemeters such as the Moving Die Rheometer (MDR) [21]. ASTM D5289 and

International Standard ISO 6502 describe the use of rotorless curemeters. In an

instrument such as the MDR, the lower die oscillates sinusoidally, applying a strain to

the rubber specimen which is contained within a sealed, pressurized cavity. The upper die

is attached to a reaction torque transducer which measures the torque response. This

design greatly improves the test sensitivity (signal-to-noise) measurements so that real

changes in a rubber compound can be detected faster. Also, because there is no rotor, the

temperature recovery of the test specimen is less than 30 seconds, compared to

approximately 4 to 5 minutes for the older ODR design. Because curing a given rubber

specimen in the MDR is closer to a true isothermal cure, the tS1 scorch time values from

the MDR are signi®cantly shorter than the tS1 scorch time values from an ODR for the

same compound at the same test conditions. At higher cure temperatures, such as 190 8C,scorch information from the MDR can be obtained in about one-half the time as from the

ODR [22]. Also, barrier ®lm can be used with rotorless curemeters which greatly reduces

clean-up time.

With the MDR, two curves are produced during cure, as shown in Fig. 2.16. The

elastic torque S0 is the traditional cure curve which is more commonly used as an

indication of cure state (the ODR produces only an S0 cure curve). The viscous torque S00

curve is a second curve generated simultaneously. Sometimes, the S00 peak can be used as

an alternate method for measuring scorch. Studies have shown that the percent drop from

this S00 peak as the cure progresses gives information regarding ®ller loadings as well as

the ultimate crosslink density [23].

A more advanced rotorless curemeter is the Rubber Process Analyzer (RPA).

Improved sensitivity to scorch is achieved by increasing the applied shear rate. When

the applied frequency or strain is increased, the resulting shear rate is increased. When a

rubber compound is tested at higher shear rates, the sensitivity to subtle scorch differences

is increased (which is also true in factory processes which apply higher shear rates to a

rubber compound). Studies have shown that higher frequencies or strains applied to a

rubber compound provide earlier warning of a scorch condition [24].


Another way to improve test sensitivity to scorch is to use the Variable Temperature

Analysis feature of the RPA. This feature allows the user to program the RPA to linearly

``ramp'' the test temperature to achieve an increase in a very controlled way from, for

example, 100 to 180 8C in 8.0 minutes. In this way, the RPA ®rst measures processing

properties at 100 8C, which is a temperature typically low enough to prevent any

vulcanization. Then the temperature is ``ramped'' up to measure scorch and cure

properties. This ``ramping of temperature'' results in better test sensitivity to scorch

differences than measuring scorch under traditional isothermal conditions [25]. In one test,

both viscoelastic processing behavior and scorch are measured with excellent precision

and test sensitivity.

2.3.4.3 Scorch by Capillary Rheometer

As discussed earlier, tests performed at higher shear rates are more sensitive to the

beginnings of scorch than tests at lower shear rates. A capillary rheometer used at a

suf®ciently high enough temperature (close to the temperature of the factory process), at a

shear rate of 300 to 1000 sÿ1, and a capillary die with a length to diameter (L/D) ratio in

excess of 10:1, has high test sensitivity to scorch. Studies have shown that capillary

rheometer scorch values relate better to computer simulation of injection molding than do

the lower shear rate ODR or Mooney scorch values [26,27]. One problem with the

capillary rheometer is that it is more complex and time-consuming than other tests, which

limits its use mainly to research and development.

2.3.5 Cure Rate

Cure rate is the speed at which a rubber compound increases in modulus (crosslink

density) at a speci®ed cure temperature or heat history. Cure time refers to the amount of

Figure 2.16 Elastic torque S0 and viscous torque S00 cure curves.


time required to reach speci®ed states of cure at a speci®ed cure temperature or heat

history. An example of cure time is the time required for a given compound to reach 50%

or 90% of the ultimate state of cure at a given temperature. Of course, determining what is

the instrumental optimum cure time for a small curemeter specimen is not the same as

determining what is the optimum cure time for a high mass, thick, rubber article cured in

the factory. This is because usually an instrumental cure is closer to an isothermal cure

(being cured at the same temperature), whereas the center portion of a thick rubber article

sees a variable temperature heat history.

The rubber technologist must take into account other factors such as the temperature

recovery time of the curemeter being used, the thickness of the rubber article cured in the

factory, the compound thermal conductivity, the variable heat history of the center of the

article, the compound cure kinetics, the compound overcure stability, etc., to determine

the optimum cure temperature and time for a rubber article in the factory. Sometimes an

empirical approach with test trials at different cure times and temperatures may be a

practical way to determine the optimum cure temperature and time for a given rubber

article. If the cure residence time is too short, then poor cured physical properties result,

especially toward the center of a thick article, because it sees less heat history. On the

other hand, if the cure residence time is too long for a thick article, then there is

deterioration in the cured physical properties, especially at the surface of the article where

it receives too much heat history. Also, a long cure time results in poor productivity. Thick

articles made from a composite of different rubber compounds represent a ``balancing

act'' by the R&D compounder in adjusting the cure properties for the different component

compounds that receive different heat histories.

There are two methods which can be used for measuring cure pro®le properties such

as cure rate and cure times:

1. rotational viscometers

2. oscillating rheometers

2.3.5.1 Cure Rate by Rotational Viscometer

As discussed earlier, Mooney Scorch with the large rotor is measured by the time

required for the viscosity to rise 5 Mooney units above the minimum (referred to as t5),

and when the small rotor is used, the scorch time is reported as the time required for

viscosity to rise 3 Mooney units above the minimum (referred to as t3). ASTM D1646

describes what is called the ``ASTM cure index.'' These indices are expressed in

Eqs. (2.12) and (2.13).

For small rotor �tS � t18 ÿ t3 �2:12�For large rotor �tL � t35 ÿ t5 �2:13�

A lower cure index means the cure is faster. However, the Mooney scorch test is not very

effective at giving cure information above t18 (for small rotor) or t35 (for large rotor)

because the rotor tears and slips at the rubber interface as the rotor rotates. The Mooney

viscometer is not effective at providing a complete cure curve because it uses a rotating

rotor, not an oscillating one. Also, the Mooney Viscometer uses large sample weights of


approximately 25 g and an unheated rotor which functions as a ``heat sink.'' Therefore, the

temperature recovery of the Mooney viscometer (the time required for the temperature of

the rubber specimen to reach the set temperature of the dies) is relatively long compared to

an oscillating rheometer such as the MDR. This temperature recovery is more critical at

higher cure temperatures.

2.3.5.2 Cure Times and Cure Rate by Oscillating Rheometer

The ODR method as described in ASTM D2084 or ISO 3417 has been used since the

1960s to measure cure times and cure rate. The tCx is simply the time to reach a given

percent x state of cure (for example tC50 is the time required to reach 50% of the state of

cure). This is illustrated in Fig. 2.17. So, mathematically, tCx is calculated as follows:

tCx is the time to produce torque �T� � �x=100� � �MH ÿML� �ML �2:14�where:

� ML � minimum torque

� MH � maximum torque

� x � percent state-of-cure

� tCx � time to a given percent (x) state-of-cure

As discussed earlier, the unheated rotor used with the ODR is functioning as a heat sink

when it is embedded in the rubber specimen. This can cause the temperature recovery time

for the rubber specimen (the time required for the temperature of the specimen to reach the

set temperature of the dies after the dies are closed) to be 4 to 5 minutes. Therefore the

cure times from the ODR are generally inaccurate in that they do not truly re¯ect the cure

times for a true isothermal cure.

Figure 2.17 Illustration of the calculation of tC50 (time to 50% state-of-cure).


With computer software used today, the maximum cure rate as well as the average

cure rate or the ASTM/ISO Cure Rate Index can easily be calculated. The maximum cure

rate is the slope of the tangent line at the in¯ection point on the cure curve while the Cure

Rate Index is calculated as follows:

Cure Rate Index � 100=�Cure Timeÿ Scorch Time� �2:15�The user of this index can select preferred cure and scorch times, such as tC90 and

tS1.

Rotorless curemeters, such as the MDR, use essentially the same methods for

calculating the cure times and cure rates as speci®ed in ASTM D5289 and ISO 6502.

However, the cure times from the rotorless curemeters are signi®cantly shorter than those

measured by the ODR for the same compound under the same set of test conditions. As

mentioned before, this is because rotorless curemeters, such as the MDR, have a cure

which is much closer to a true isothermal cure. The temperature recovery for the MDR

[28] is only 30 seconds and the temperature drop when loading a sample is much less

than for an ODR. Therefore, cure kinetic studies based on the German Standard DIN

53529 can be performed much better with the MDR [29]. If the ODR were used, there

would be a high degree of error because of the ODR's much longer temperature

recovery.

While the MDR is curing a compound at near-isothermal conditions (the same

temperature over time), thick rubber article cures are not isothermal. The center of a

thick section of a rubber article cured in a mold is exposed to a variable heat history. A

thermocouple in the center of a rubber article can measure the rise in temperature as the

rubber article is cured. This time-temperature pro®le can be used by the RPA software to

give the same heat history to the compound as measured by the thermocouple. Because the

RPA has low mass dies and very ef®cient heaters, the temperature of the upper and lower

dies can exactly match the required time-temperature pro®le. The RPA can also match a

drop in temperature after demolding through the use of a forced air cooling system. This

Variable Temperature Cure technique is an empirical way to estimate state-of-cure for the

center of a thick section. It should be understood that the cool down portion of the test can

introduce an error, i.e., the S0 can increase from both crosslinking and the drop in

temperature.

2.3.6 Ultimate State of Cure

Ultimate state of cure refers to the ``ultimate crosslink density.'' Strictly speaking, the

``best'' state of cure can be quite different for one rubber property, such as tear resistance,

compared to another rubber property, such as rebound. When discussing processability,

however, the ultimate state of cure is usually measured as the maximum elastic torque

(MH). This is not a perfect method of measuring the ultimate state of cure, but it is very

practical. One problem with using maximum torque to measure ultimate state-of-cure

occurs when the compound displays a ``marching modulus.'' This phenomenon occurs

when the S0 torque never reaches a plateau during cure. In this case, an arbitrary cure time

must be set to measure the maximum S0 torque response. This selected cure time should be

located where the rate of S0 increase has slowed signi®cantly.


There are two tests used in production to measure the ultimate state of cure:

1. ring testing

2. oscillating rheometer

2.3.6.1 Ring Testing

Ring testing can be used in a production setting provided the ring specimens can be cured

fast enough. This method is discussed in Chapter 3 as ASTM D412 and ISO 37 for tensile

testing.

2.3.6.2 Oscillating Rheometer

The most common and practical test to measure the ``ultimate state of cure'' is by using

the maximum S0 torque (MH) from the oscillating rheometer. The ODR (ASTM D2084 and

ISO 3417) is commonly used to measure MH. The MDR can also measure MH (Ref. ASTM

D5289); however it can measure the viscous torque S00 @ MH, as well. This is illustrated in

Fig. 2.18. As can be seen, the S00 response sometimes changes more from carbon black and

oil variations than the S0 response. In addition, the RPA can give not only an S0 and S00 curecurve, but it can also quickly drop the die cavity temperature and measure after-cure

dynamic properties at a designated lower temperature. These properties often relate better

to product performance. An example of an RPA after-cure temperature sweep with the

tan d responses, is given in Fig. 2.19. In this example, the differences in the cured tan dvalues are much greater at the lower temperature.

2.3.7 Reversion Resistance

Reversion resistance is the resistance of a rubber compound to deterioration in vulcanizate

properties usually as a result of extended curing times. This property is particularly

Figure 2.18 S0 and S00 cure curves for selected vibration isolator compounds.


important when a rubber part or compound experiences too much heat history during

curing. For example, the outside surface of a thick article may be exposed to an excessive

heat history, causing reversion to occur. Rubber compounds based on natural rubber

commonly revert at higher cure temperatures.

Oscillating rheometers are the most practical method for measuring reversion. The

ODR, MDR, and RPA can measure the drop in the S0 response after MH. When the S0

elastic torque peaks and then drops because of reversion, the S00 torque response and the

tan d both rise with reversion. There is some evidence that these dynamic properties are

more sensitive to reversion [30]. The RPA can increase the test sensitivity to reversion

even more by performing an in situ post cure aging test at an elevated temperature such as

190 8C and then measuring the percent change in the cured tan d at a lower temperature

such as 60 8C [31].

2.3.8 Green Strength

Green strength is the tensile strength and/or tensile modulus of an uncured rubber

compound. This important processing property relates to the compound's performance

in extrusions, calendering, and conveyor belt or tire building, particularly the second-stage

building machine for radial tires. If tires are constructed with rubber compounds with poor

green strength, they may fail to hold air during normal expansion in the second stage of

tire building prior to cure. High molecular weight, strain crystallizing elastomers (such as

natural rubber) tend to exhibit good green strength.

The only standardized test for measuring green strength is International Standard ISO

9026, which speci®es the preparation of dumbbells for tensile testing. This method calls

for testing ®ve dumbbell specimens and reporting the medians and ranges found.

Research done in 1996 showed that high strain testing with the RPA could correlate to

green strength for a series of natural rubber compounds [32].

Figure 2.19 RPA after cure temperature sweep for two different tire tread stocks.


2.3.9 Tackiness

Tackiness refers to the ability of an uncured rubber compound to stick to itself or another

compound with a short dwell time and a moderate amount of applied pressure [33]. This

property is very important whenever rubber products, such as tires or conveyor belts, are

built by laying one calendered or extruded rubber ply on top of another. The uncured

product must hold together before it is placed in a mold or press for cure. Usually

compounds based on natural rubber have good building tack. Compounds based on EPDM,

on the other hand, usually have poor building tack. Many times, tacki®ers are added to a

compound to improve tack.

Rubber compounds consist of many ingredients with differing degrees of solubility.

Some of these ingredients may separate from the compound under certain cooling

conditions, exuding to the compound's surface to impart a surface bloom. Many times,

this bloom destroys building tack. Some compounding ingredients which may bloom are

sulfur, accelerators, antidegradants, petroleum oils, zinc stearate, and waxes.

There are no ASTM or ISO test standards for measuring the tack of rubber

compounds. However, the most widely used tack testing instrument is the Tel-Tak

Tackmeter, introduced by Monsanto in 1969 [34].

2.3.10 Stickiness

Stickiness refers to the nature of a rubber compound to stick to non-rubber surfaces such as

metals or textiles. Too much stickiness to metal surfaces can result in poor mill and

calender release and problems in other processing equipment. However, not enough

stickiness can result in rubber compound slippage against metallic surfaces in extruders

or internal mixer rotors. Adjusting temperature can sometimes control stickiness. Certain

compounding ingredients such as external lubricants or release agents are sometimes used

to control the stickiness level. These agents, however, should be used with caution because

adhesion and other compounding properties might be affected.

There are no ASTM or ISO test methods developed to measure stickiness. However,

the Tel-Tak Tackmeter developed by Monsanto in 1969 can be used to measure not only

compound tackiness, but stickiness to a stainless steel surface as well [35].

2.3.11 Dispersion

Dispersion is a property which de®nes how well ®ller aggregates and particles are

dispersed in a rubber compound from a mixing process. This property relates not only

to percent carbon black dispersion, but also to non-black ®llers such as clays, silica,

titanium dioxide, calcium carbonate, etc. Rubber curatives, such as accelerators and sulfur,

can also be poorly dispersed. Poor dispersion can sometimes be a particularly critical

characteristic with curatives because they are commonly added late in the mixing cycle.

Poor dispersion for a rubber batch can result in poor stock uniformity and highly variable

cured physical properties such as ultimate tensile strength. It is well known that poor

dispersion can decrease abrasion, tear, and fatigue resistance as well as hurt ¯exometer


heat buildup and other dynamic properties. Test method ASTM D2663 lists three different

methods for quantifying percent carbon black dispersion [36].

2.3.12 Stock Storage Stability

Stock storage stability is de®ned as the period of time a given mixed stock can be stored on

the factory ¯oor and remain usable. Stock can be affected by scorch time [37]. Usually, but

not always, the longer a mixed stock is stored, the lower its scorch safety. With storage,

rheological properties change as a result of an increase in interaction between the rubber

and carbon black (called bound rubber). This can be seen by a rise in compound's Mooney

viscosity or oscillating rheometer minimum torque ML (see Section 2.3.1). LeBlanc and

Staelraeve reported on the advantages of large strain RPA tests for improved sensitivity to

storage maturation [38]. Not only mixed stocks, but raw elastomers, such as natural rubber,

can manifest changes in their rheological properties with storage. For example, the storage

of natural rubber can produce a rise in viscosity and storage hardening [39]. The RPA can

also be used to measure the storage hardening properties of natural rubber [40].

2.3.13 Mis-Compounding

Mis-compounding occurs when an error is made in the labeling or weighing of compounding

ingredients before mixing or when not all of the ingredients are properly loaded into the mixer.

Establishing speci®cations for batch weight is one check for compounding ingredient

weighing errors. Many processing tests are affected by such errors. The careful monitoring

of dynamic property variations and pattern changes with the MDR or the RPA can help

determine which compounding ingredient might have been weighed in error [41,42]. Also, the

Compressed Volume Densimeter described in ASTM D297 can easily measure the density of

unvulcanized rubber compounds. If, for example, the compound density did not change, but

the dynamic properties from the MDR or RPA suggested a change in the ®llers, this may

indicate that the wrong type of carbon black was used.

2.3.14 Cellular Rubber Blow Reaction

This chemical reaction occurs as a result of the decomposition of one or more blowing

agents which generate gas during the cure. The generation of this gas is necessary to

produce a cellular rubber product (see Chapter 21). To achieve the proper cellular

structure, the cure reaction and blow reaction must be in balance. Therefore, it is

important to measure the blow reaction.

Currently there are no standards available for tests to monitor blow reactions.

However, instruments, such as the MDR-P [43] from Alpha Technologies, measure both

the cure reaction and the blow reaction simultaneously. This is important because if these

two reactions are not balanced, then unacceptable cell size or structure may result.

Parameters somewhat analogous to what is used to describe the cure reaction are used

for the blow reaction. For example, the minimum and maximum pressure as well as time to

50% of the maximum pressure are calculated automatically [44,45].


2 Compound Processing Characteristics and Testing2 Compound Processing Characteristics and Testing...

Documents

Transcript of 2 Compound Processing Characteristics and Testing2 Compound Processing Characteristics and Testing...